The video surveillance camera manufacturer Hanwha Vision announces that it’s achieved ISO/IEC 42001 certification, the international standard for a certifiable Artificial Intelligence Management System (AIMS). The company describes it as a formal promise to partners and customers that its entire process for developing and deploying AI is governed by a ‘human-centric’ philosophy. The company writes:
The global community is moving fast to ensure AI remains a force for good. With the enforcement of the EU AI Act, the world’s first comprehensive AI law, and similar legislative movements in the US and Asia, the regulatory grey area is disappearing.
These regulations are not just bureaucratic hurdles. They are a response to genuine concerns regarding data privacy, algorithmic bias, and the potential for surveillance overreach. For an end-user, choosing a provider that aligns with these global standards is no longer just a matter of ethics. It is a matter of long-term operational security. By adhering to ISO/IEC 42001, Hanwha Vision proactively meets these stringent requirements, ensuring our customers are protected from the legal and ethical risks of non-compliant AI.
To maintain this high standard of excellence, we have institutionalised the Responsible AI (RAI) Council. This dedicated decision-making body sets internal compliance standards, conducts pre-release risk assessments, and monitors ongoing AI performance. By doing so, they ensure every innovation is rooted in our core Ethical AI Principles.
- Safety and Reliability: We believe AI must be a dependable partner, performing consistently in any environment to ensure public safety.
- Privacy and Human Dignity: We respect individual rights as a non-negotiable value, ensuring technology never compromises the dignity of the people it serves.
- Transparency and Fairness: We are committed to integrity, rejecting ‘black box’ processes to ensure our intelligence is both auditable and impartial.
Principles to life
How do these principles translate into the technology inside a security system? At Hanwha Vision, our AI management framework serves as the strict technical requirements that guide how our AI is built and deployed.
- Privacy by Design: We use AI to protect privacy, not just to monitor. Our systems can mask faces or sensitive areas in real-time, ensuring that security data is used only for its intended purpose without compromising individual anonymity.
- Eliminating Algorithmic Bias: A security system must be fair. We rigorously audit our training datasets to ensure our AI recognises people and objects accurately across diverse environments and demographics, preventing discriminatory outcomes.
- Data Integrity and Security: AI is only as good as the data it learns from. We strictly manage our datasets to prevent any unauthorised tampering or data poisoning. By securing both the data and the algorithms, we ensure our AI provides consistent, accurate, and untampered insights that our customers can always trust.
- Transparency and Explainability: We believe users should understand how AI decisions are made. Our framework ensures that our AI processes are auditable and transparent, moving away from the opaque black box approach to technology.
The video surveillance industry is entering a new era where technical specs alone no longer define excellence. As AI becomes more autonomous, the true competitive edge lies in accountability.
